sign in sign up
<<
Home , Alderney, Bailiwick, Bailiwick of Guernsey, British Isles, Guernsey

222Journal ofNeurology, Neurosurgery, and Psychiatry 1995;58:22-26

Multiple sclerosis in island populations: J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.22 on 1 January 1995. Downloaded from prevalence in the of and

G Sharpe, S E Price, A Last, R J Thompson

Abstract ous neurological disability in young adults; it The aim of this study was to establish for affects some 60 000 people in the United the first time the prevalence of multiple Kingdom and perhaps two million people sclerosis in the Bailiwicks of Guernsey worldwide.' The disease shows an unusual and Jersey, as representing the most geographical distribution in becoming com- southerly part of the Isles. All moner with increasing distance from the patients with multiple sclerosis in the Equator in both the northern and southern resident on prevalence hemispheres.2 There is evidence for a similar day were identified by contacting all geographical gradient in the .2 medical practices, Multiple Sclerosis, Different studies have reported multiple scle- and Action Research for Multiple rosis to be 1 9 to 3 1 times commoner in Sclerosis societies by letter and visits. women than men and to have a peak age of The crude overall prevalence rates were onset of about 30 , being rare in child- 1131100 000 (95% confidence interval hood and after the age of 50.2 The clinical fea- (95% CI) 90-3-135.7) and 86-7/100 000 tures, sex ratio, and age specific incidence (95% CI 63.3-110.0) for the Bailiwicks of curves for multiple sclerosis are similar what- Jersey and Guernsey respectively. When ever the underlying frequency of disease, sug- standardised to the age and sex structure gesting the same worldwide aetiology.2 of a previously reported Northern Among the many epidemiological surveys population the standardised prevalence of multiple sclerosis in the past 40 years, sev- ratios were 120-21100 000 (95% CI eral intriguing reports have studied the preva- 96-0-144-3) for Jersey and 95 61100 000 lence of the disease in geographically defined (95% CI 69-9-121.3) for the of island populations. The and Guernsey. When compared with recent islands have the highest prevalence of the dis- studies in the northern ease ever recorded-namely, 309 and 184 per the prevalence rates for multiple sclerosis 100 000 respectively.' The prevalence on in the Channel Islands lend some support Sardinia4 and Sicily5 has been reported to be to the proposed latitudinal gradient in the over 45 per 100 000, whereas on nearby British Isles although the standardised the prevalence is tenfold lower.6 prevalence ratio in the Bailiwick ofJersey Clustering of cases of multiple sclerosis has is similar to those found in recent studies been recorded on Orkney7 and on the small http://jnnp.bmj.com/ of southern . The standardised island of Key West off the coast of Florida.8 prevalence rates of probable and definite An "epidemic" of multiple sclerosis in the multiple sclerosis for the male popula- Faroes after occupation by British troops has tions were 37 31100 000 (95% CI 17-9-56.7) been proposed,9 and a similar rise in preva- for the and lence after troop incursions has been claimed 45 51100 000 (95% CI 26.3-64.7) for the on Orkney,'0 ," and Sardinia.'2 That

Bailiwick ofJersey whereas the standard- the incursions caused the apparent increase in on September 26, 2021 by guest. Protected copyright. ised prevalence rates for the female popu- the prevalence of multiple sclerosis in these lations were 97 51100 000 (95% CI situations has in most cases been disputed.2 73-9-143.5) and 139*5/100 000 (95% CI The Channel Islands lie 70-100 miles 112-6-181-2) respectively. Thus there is a south of the coast of between 490 striking and unexplained 43% higher and 500 latitude and about 10-30 miles west prevalence of probable and definite mul- of the French coast, with the nearest island University Clinical tiple sclerosis in the female population of () only eight miles from the Biochemistry, Level Jersey compared with that of the . The British Channel D, South Laboratory Block, Southampton Bailiwick of Guernsey. This seems to be Islands are not part of the United Kingdom, General Hospital, due to an unusually low prevalence of the but are allied to the Tremona Road, disease among the female population of English crown since the Norman invasion. Southampton S09 4XY, UK the Bailiwick of Guernsey compared with They represent the most southerly part of the G Sharpe that of the United Kingdom . British Isles. These Islands were invaded by S E Price in and July 1940 and liberated A Last (3 Neurol Neurosurg Psychiatry 1995;58:22-26) in 194513 and were the only part of the R J Thompson British Isles occupied in the second world Correspondence to: Professor R J Thompson. war. The islands are divided into the Received 11 February Keywords: multiple sclerosis; prevalence survey Bailiwicks of Guernsey and Jersey. The 1994 and in revised form 3 May 1994. Bailiwick of Guernsey consists of Guernsey, Accepted 13 May 1994 Multiple sclerosis is a common cause of seri- Alderney, , and other small islands, and Multiple sclerosis in island populations: prevalence in the Bailiwicks of Guernsey andJ7ersey 23

the Bailiwick of Jersey contains several small ple sclerosis was then subjected to careful J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.22 on 1 January 1995. Downloaded from uninhabited islands with only Jersey itself investigation. Sufficient clinical details to populated. Health care services in the two allow classification by both the criteria of Bailiwicks are independent of the NHS and of Poser et al 19 and of Allison and Millar'0 were each other. Also, as a popular holiday destina- sought. This was obtained either by three tion and an attractive residential choice due to authors (GS, SEP, AL) being allowed access favourable tax concessions, the Channel to the patients' notes or by the medical practi- Islands are not as geographically isolated as tioners filling in a detailed questionnaire. The some other islands-for example, Orkney and hospital notes of patients who had attended Shetland.7 No previous survey of the preva- the Wessex Neurological Centre were also lence of multiple sclerosis in the Channel studied. Local Multiple Sclerosis Societies Islands has been carried out. The present were requested to ask their members to supply study was undertaken to determine the preva- details of place of birth and age of onset of the lence of multiple sclerosis in the two disease. Forty eight patients on Jersey and 36 Bailiwicks in the light of recent estimates of on Guernsey and Alderney who volunteered the prevalence of multiple sclerosis in south- had a venous blood sample taken by a ern Britain,'4-18 and the proposed north-south member of the team. gradient in the prevalence ofmultiple sclerosis in the British Isles.' STATISTICAL ANALYSIS Population statistics were obtained from the census taken on prevalence day on Jersey,21 Methods Guernsey,22 and Alderney." No formal census CASE ASCERTAINMENT data are gathered on Sark and it was esti- A list of potential patients with multiple scle- mated that 560 people were resident there: we rosis was compiled with information obtained did not find a single case of multiple sclerosis from several sources, as the Southampton in this population and have excluded it from study'6 had shown the value of this approach. further analysis. To allow for differences The medical practitioners from both between islands in age and sex structure and Bailiwicks were a major source ofinformation. to allow comparison with published work, all We also received help from the Multiple crude prevalence data have been standardised Sclerosis and Action Research for Multiple against the 1961 popula- Sclerosis (ARMS) Societies in both sets of tion24 by two different methods. The indirect islands, and also from the hyperbaric oxygen method was used to calculate standardised unit on Guernsey. The nearest British major prevalence ratios with 95% confidence inter- referral centre for neurology is the Wessex vals (95% CIs) for each Bailiwick. To test the Neurological Centre at Southampton; this has significance of differences in prevalence become the predominant specialist referral between the Islands and the 1961 Northern centre for the Channel Islands but many Ireland population for specific age groups, it patients have been seen elsewhere. Also, a was assumed that the observed number of consultant neurologist from Southampton cases follows a Poisson distribution with a visits Jersey regularly. Consultant neurologists mean equal to the expected number of cases

at the Wessex Neurological Centre were under the standard prevalence rates. The http://jnnp.bmj.com/ therefore approached. direct method was used to calculate age stan- From these sources we compiled a provi- dardised prevalence rates for each Bailiwick sional list of potential cases of multiple sclero- and the difference in standardised rates sis covering the Bailiwicks of Guernsey and between islands, along with 95% CIs.25 Jersey. This was then used for a more detailed study. Medical practitioners were contacted again by letter and requested to provide Results details (name, date of birth, sex, and whether The provisional register of potential patients on September 26, 2021 by guest. Protected copyright. the patient was aware of the diagnosis) of any with multiple sclerosis in the Channel Islands patients under their care with multiple sclero- contained 193 names; of these 45 were even- sis who were alive and resident on the tually excluded because they were not resi- Channel Islands on prevalence day. dent on prevalence day (n = 8), or because After this, consent was sought to approach the diagnosis of multiple sclerosis could not the patient and obtain a venous blood sample be confirmed (n = 9), or the person was not for a separate genetic study. Prevalence days traced (n = 15). Additionally, one patient was were 10 1991 for the Bailiwick of registered with more than one practice and six Jersey and 21 April 1991 for the Bailiwick of were registered under both married and Guernsey. These days were chosen to coin- maiden names. Of the 45 patients 21 came cide with local census days on each set of from Jersey and 24 from Guernsey. A diagno- islands. (The difference between the two sis of multiple sclerosis was accepted if it was prevalence days is not considered relevant to confirmed by information from the patient's the results of the study.) Further letters were medical practitioner or from the medical sent to doctors not replying to initial enquiries notes. Some excluded patients had other diag- and non-responders were contacted by phone, noses-for example, hereditary spastic para- or personal visit, or both. paresis (which can mimic multiple sclerosis) (n = 4), Parkinson's disease (n = 1), or CLASSIFICATION OF PATIENTS benign intracranial hypertension (n = 1). The provisional list of possible cases of multi- Others had suspicious symptoms but no signs 24 Sharpe, Price, Last, Thompson J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.22 on 1 January 1995. Downloaded from (n = 4). Every medical practice in the lences and the standardised prevalence ratios Bailiwicks of Guernsey and Jersey responded with the populations of the two Bailiwicks age to this survey. Table 1 shows the number of and sex standardised against the 1961 resident cases found in each Bailiwick on Northern Ireland population.24 Also shown prevalence day together with the recorded are the figures from other recent studies in the populations from the official census returns. It south of the United Kingdom.'4-8 The census also shows the 53 patients from the Bailiwick reports showed that the age and sex structures of Guernsey and the 95 patients from Jersey of the populations of the two Bailiwicks were classified according to the criteria of Allison almost identical (results not presented). and Millar20 and those of Poser et al. 9 Ninety Further analysis was confined to patients with four per cent of the patients of the Bailiwick of multiple sclerosis who came into the "definite Guernsey and 88% of the patients on Jersey and probable" categories of the criteria of came under the "probable and possible" Poser et al.'9 These comprised 85% of the Allison and Millar classification20; 85% of the patients from Guernsey and Aldemey and patients of the Bailiwick of Guernsey popula- 88% of the patients from Jersey. The figure tion and 88% of the Jersey population were shows the age specific prevalence of probable fully classifiable under the system of Poser et and definite cases of multiple sclerosis for al.'9 Of those who could not be readily classi- both sets of islands compared with that of the fied under the Poser system 10 could not be Northem Ireland population.'4 Whereas the categorised even with full clinical details overall shape of the distribution is similar in (mainly those with progressive spinal cord dis- all three cases (especially at earlier ages), both ease) and with the remainder (n = 9) we were Bailiwicks have higher prevalences in older unable to attain sufficient clinical details to age groups than the original Northem Ireland make classification possible. The four patients study.20 At ages 60 and older, 8-35 cases from Alderney in the Bailiwick of Guernsey would have been expected in Jersey by apply- (one male and three female) were all in the ing the prevalence of the standard popula- clinically definite category. Of the total identi- tions, but 18 cases were found. This excess fied cases of multiple sclerosis, 19 (13%) had a was highly significant (p = 0003). At these positive family history but in only 5% of cases age groups in Guemsey and Aldemey, 12 was a first degree relative also affected. Only cases were observed compared with 6-95 one affected sib pair was found on Guernsey, expected (test for excess p = 0 5). Table 3 and one patient with a normal twin was resi- shows the crude prevalence rates and the age dent on Jersey, as was one male patient with a adjusted rates for male and female cases in the sister not resident in the Channel Islands but Channel Islands. Surprisingly, whereas the being investigated for the disease elsewhere. male populations only showed a difference of Table 2 shows the overall crude preva- 8-2 cases per 100 000 males between the two

Table 1 Clinical classification ofpatients with multiple sclerosis on the Channel Islands _- Jersey,1991 ---- Guernsey and Alderney,1991 Bailiwick Northern Ireland,1961

Guernsey J7ersey 250 http://jnnp.bmj.com/ (Population 61 624) (Population 84 082)

Total cases 0>. 200- recorded 53 95 Female:male sex ratio 2-6:1 3-1 4X 150

Alison and Millar criteria20: 0 Probable 42 78 0 Possible 8 6 0 o

Early 3 11 0 on September 26, 2021 by guest. Protected copyright. Poser et al criteria'9: CDMDS 40 80 X. 50 LSDMS 1 1 U) CPMS 2 1 LSPMS 2 2 0 Unclassifiable 8 11 0 15 30 45 60 75 75+ Age (y) CDMS = Clinically definite multiple sclerosis; LSDMS = laboratory supported definite multiple sclerosis; CPMS = Clinically probable multiple sclerosis; LSPMS = Laboratory Age specific prevalence ofprobable and definite cases of supported probable multiple sclerosis; Unclassifiable = multiple sclerosis (MS). Unclassifiable with Poser et al criteria.

Table 2 Comparison ofChannel Islands study with other recent south British studies Guemsey and Southampton and SE SE Jersey 1991 Alderneyt 1991 Cambridge 1990 SWHHA 1987 (Sutton) 1985 1985 No of cases 95 53 374 411 195 441 Population 84 082 61 164 288 410 411 000 169 600 376 710 Crude prevalence rates/100 000 113 (90-3-135-7) 86-7 (63 3-110 0) 130 (117-143) 99 (89-109) 115 (99-131) 117 (106-128) *SPR 120-2 (96-0-144-3) 95-6 (69-9-121-3) 121 (108-133) 115 (108-133) 129 (111-127) 139 (126-150) *Standardised prevalence ratio. The figure shown is the one that would have been obtained if these populations had the same age and sex structure as the popula- tion surveyed in Northern Ireland 196120 and taking the prevalence found in that study as = 100."4 tThe Bailiwick of Guernsey consists of Guernsey, Alderney, and Sark. The population ofAlderney was 2297 with four recorded cases (one male, three female). No cases were recorded on Sark (population estimate 560) which was excluded from further analysis. SWHHA = South West Health Authority. Multiple sclerosis in island populations: prevalence in the Bailiwicks of Guernsey andJersey 25 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.22 on 1 January 1995. Downloaded from Table 3 Age standardised prevalence rates ofprobable and definite multiple sclerosis in prevalence of multiple sclerosis there is of male andfemale populations in the Channel Islands value in deciding whether a latitudinal gradi- Age standardised* ent in the prevalence of the disease exists in Crude prevalence prevalence the British Isles as seems to occur on a world- Males: wide basis.2 Whereas the current age and sex Bailiwick of Guernsey 47-6/100 000 37 3/100 000 (95% CI 17 9-56 7) standardised prevalence ratio of 120X2 (if 100 Bailiwick ofJersey 53 8/100 000 45-5/100 000 = Northern Ireland 1961) in the Bailiwick of (95% CI 26 3-64 7) Jersey is little different from the standard Females: prevalence ratios in Southampton (115),16 Bailiwick of Guernsey 110 2/100 000 97-5/100 000 (95% CI 73-9-143-5) Cambridge (121),'8 and Sutton (129),'4 it is Bailiwick ofJersey 162-0/100 000 139-5/100 000 considerably lower than the most recent sur- (95% CI 1126-181-2) vey of north east (221).26 The *To Northern Ireland 1961 crude prevalence = 65 8/100 000 (male) 94-6/100 000 (female). Bailiwick of Guernsey has an even lower Difference in age standardised prevalence between two Bailiwuicks: male = 8-2 cases/100 000. prevalence ratio of 95-6. (19-8 to 36-1), p value for difference = 0-6; female = 42 cases/100 000 -6-4 to 90 3) p = 0-09. These figures overall would therefore lend at least some support to the proposed latitudinal gradient in preva- Bailiwicks (95% CI 19-8-36-1, p = 0 6), the lence of multiple sclerosis in the British Isles,2 female populations showed a profound differ- at least between northern and southern ence, with a large excess of 42 cases per Britain. Between the two Bailiwicks the age 100 000 of the female population seen on specific prevalence profiles ofmultiple sclerosis Jersey (95% CI -6-4 to 90 3, p = 0 09). The for probable or definite cases show signifi- mean age of male patients with multiple scle- cantly higher prevalence in older age groups rosis was 49-6 for the Bailiwick of Guernsey (60+) on both sets of islands (Jersey, p = and 53-6 for Jersey and the corresponding 0-003; Guernsey, p = 0-5) than in the mean ages for female patients were 46 1 and Northern Ireland population, which presum- 43-8 respectively. We were able to establish ably reflects improved longevity of patients in the of onset of the disease in 46 (87%) of the 30 years since that study was performed.24 the patients on Guernsey and Alderney and in The familial incidence of multiple sclerosis in 69 (73%) of the patients on Jersey. There was the Channel Islands also seems to be lower no evidence of clustering by year of onset than in some surveys2 and only seven (5%) among these patients, as claimed in a previous patients have a first degree relative also Orkney study.7 We were unable to find reli- affected. able information on the prevalence of multiple The striking finding in the present study is sclerosis in the Channel Islands before the the unexpectedly high prevalence of female second world war and in only one patient in patients with probable or definite multiple this study was the onset of the disease before sclerosis on Jersey compared with the 1940. Among the patients in whom age of Bailiwick of Guernsey (139-5/100 000 v onset was established there was a tendency for 97-5/100 000), which largely accounts for the it to be earlier in the female (but not male) difference in overall prevalence between the patients on Guernsey than in those on Jersey. two populations (table 3). The female:male We were able to establish the place of birth in ratios of the general population of the two 40 (75%) of the 53 patients in the Bailiwick of Bailiwicks are almost identical (1:1 *07 and http://jnnp.bmj.com/ Guernsey and in 57 (60%) of the 95 patients 1:1 05). Because the methods of ascertain- on Jersey. Of these, 70% of the patients on ment were the same on both sets of islands Guernsey and Alderney and 58% of the and because only probable or definite cases patients on Jersey were native to their respec- were included, this difference seems to be tive islands. From the census returns21-23 65% real. The observed prevalence per 100 000 of the total population of Guernsey and females on Jersey is in line with that seen in Alderney were born on the islands with 77% recent surveys of southern Britain.' 148 Thus of the remaining 35% of population originat- the female prevalence rate in the Bailiwick of on September 26, 2021 by guest. Protected copyright. ing elsewhere in the British Isles. On Jersey Guernsey should be regarded as unusually 52% of the total population are native with low compared with mainland Britain. There is 76% of the remaining 48% coming from else- a dearth of prevalence data for nearby , where in the British Isles. although a recent paper has quoted a figure of 40/100 000.27 We are unaware of any estab- lished difference in the genetic make up of the Discussion populations of the two Bailiwicks and it is dif- The present study is the first survey of the ficult to see why such factors should only prevalence of multiple sclerosis in the apply to the female and not the male popula- Channel Islands. The methods used were tion. similar to other recent studies in southern The first two authors made equal contributions to this work. England.'6 18 Because of the geographically We are grateful to the medical practitioners, the officers and circumscribed members of the Multiple Sclerosis and ARMS Societies, and island nature of the popula- especially the patients in both the Bailiwicks of Guernsey and tions studied, the many sources of case identi- Jersey for taking part in this study. We also acknowledge the help of the neurologists at the Wessex Neurological Centre. fication, and the high response rate to We thank the for providing the Third inquiries, we believe that the estimate of num- Guernsey Fellowship (GS) via the Wessex Medical Trust. bers of patients in This work was also supported by the Multiple Sclerosis the Bailiwicks of of and Northern Ireland. SEP is an and Jersey is accurate. As the Channel Islands MRC Training . A Shiel (MRC Epidemiology Unit) is lie 70-100 miles thanked for expert statistical advice. F Lander, K Debrah, J south of the United Dean, B Morgan, and S Linham are thanked for help at various Kingdom mainland, information on the stages of this project. 26 Sharpe, Price, Last, Thompson

1 Compston DAS. The dissemination of multiple sclerosis. 13 King P. The Channel Islands at war. London: Robert Hale, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.58.1.22 on 1 January 1995. Downloaded from Journal of the Royal College ofPhysicians of London 1990; 1991. 24:207-18. 14 Williams ES, McKeran RO. Prevalence of multiple sclero- 2 Martyn CN. The epidemiology of multiple sclerosis. In: sis in a South London . BMY 1986;293:237-9. Matthews WP, ed. McAlpine's multiple sclerosis. London: 15 Lockyer MJ. Prevalence of multiple sclerosis in five rural Churchill Livingstone, 1991:3-212. Suffolk practices. BMJ 1991;303:347-8. 3 Poskanzer DC, Prenney LB, Sheridan JL, Kandy JY. 16 Roberts MHW, Martin JP, McLellan DL, et al. The preva- Multiple sclerosis in the Orkney and Shetland Islands. lence of multiple sclerosis in the Southampton and Y Epidemiol Community Health 1980;34:229-39. South West Hampshire Health Authority. J Neurol 4 Rosali G, Aiello A, Pirastow MI, et al. Sardinia, a high risk Neurosurg Psychiatry. 1991;54:55-9. area for multiple sclerosis: a prevalence and incidence 17 Swingler RJ, Compston DAS. The prevalence of multiple study in the of Alghero. Ann Neurol 1987;21: sclerosis in South East Wales., J Neurol Neurosurg 190-4. Psychiatry 1988;51:1520-4. 5 Dean G, Grimaldi G, Kelly R, Karhausen L. Multiple 18 Mumford CJ, Fraser MB, Wood NW, Compston DAS. sclerosis in southern I: Prevalence in in Multiple sclerosis in the Cambridge health district in 1975. Y Epidemiol Community Health 1979;33:107-10. East Anglia. J Neurol Neurosurg Psychiatry 1992;55: 6 Vassallo L, Elian M, Dean G. Multiple sclerosis in 877-82. II. prevalence in Malta 1978. 19 Poser CM, Pato DW, Scheinberg W, et al. New diagnostic _JEpidemiol Community Health 1979;33:110-3. criteria for multiple sclerosis: guidelines for research pro- 7 Poskanzer DC, Walker AM, Prenny LB, Sheridan JIL. The tocols. Ann Neurol 1983;13:227-31. aetiology of multiple sclerosis in temporal-spatial clus- 20 Allison RS, Millar JHD. Prevalence and familial incidence tering indicating two environmental exposures before of disseminated sclerosis (A report to the Northern onset. Neurology 1981;31:708-13. Ireland Hospitals Authority on the results of a three year 8 Sheremata WA, Poskanzer DC, Withum DG, MacLeod survey). MedJ 1954;23(suppl 2):5-49. CL, Whiteside ME. Unusual occurrence on a tropical 21 Jersey census 1991, States ofJersey. island of multiple sclerosis. Lancet 1985;ii:618. 22 Guernsey census 1991, States of Guernsey. 9 Kurtzke J, Hullested K. Multiple sclerosis in the Faroe 23 Alderney census 1991, States of Guernsey. Islands III: An alternative assessment of the three 24 Millar JHD. Multiple sclerosis in Northern Ireland. epidemics. Acta Neurol Scand 1987;76:317-39. In: Clifford Rose, F. ed. Clinical neuroepidemiology. 10 Martin JR. Troop-related multiple sclerosis outbreak in Tunbridge Wells: Pitman Medical Ltd, 1980: 222-7. the Orkneys. J Epidemiol Community Health 1987;41: 25 Armitage P. Statistical methods in medical research. Oxford: 183-4. Blackwell, 1971. 11 Kurtze JF, Gudmusson KR, Bergmann S. Multiple sclero- 26 Phadke JG, Downie AW. Epidemiology of multiple sclero- sis in Iceland: Evidence of a post-war epidemic. sis in the north-east (Grampian ) of Scotland-an Neurology 1982;32: 143-50. update. J Epidemiol Community Health 1987;41:5-13. 12 Rosati G, Aiello I, Granieri E, et al. Incidence of multiple 27 French Research Group on Multiple Sclerosis. Multiple sclerosis in Macower, Sardinia 1912-1981; onset of the sclerosis in 54 twinships-concordance rate is indepen- disease after 1950. Neurology, 1986;36:14-9. dent of zygosity. Ann Neurol 1992;32:724-7. http://jnnp.bmj.com/ on September 26, 2021 by guest. Protected copyright.